INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ÂłÂłÂłÂłÂłÂłÂłÂł

-0.72

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

-0.71

 Router

-0.66

oult

-0.66

oldown

-0.63

ulhu

-0.62

eele

-0.62

Stra

-0.62

 Balkans

-0.61

estern

-0.61

POSITIVE LOGITS

zeb

0.67

furt

0.66

dx

0.65

 Mell

0.64

 Hawking

0.61

 unreasonable

0.60

 defective

0.60

 Chrys

0.59

objects

0.58

 engagements

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.